Can eXpress be used for finding differentially expressed genes in RNA-Seq analysis?
eXpress does not test for differential expression but the (rounded) "effective counts" in the results.xprs file can be used in downstream count-based analysis tools such as DEGSeq, edgeR, baySeq, or DESeq. However, none of these tools are suitable for isoform level differential expression and until our own differential expression tool is completed, we currently recommend Cuffdiff for that purpose.
Can eXpress be used to perform genome or transcriptome assembly?
No, eXpress is solely a quantification tool and cannot be used for assembly. However we believe the streaming algorithms underlying eXpress can be adapted for use in assembly as well and we are currently exploring ideas for low-memory streaming based assembly. For now we recommend Cufflinks for reference-based transcriptome assembly. Trinity, Oases, and Trans-ABySS can be used for de novo transcriptome assembly. There are numerous new short-read assembly tools that can be used for DNA assembly.
Is eXpress restricted to RNA-Seq analysis?
The methods in eXpress are general and can be used in any application where it is necessary to quantify the relative abundances of sequences that have been sampled using short reads. RNA-Seq is the application that we have focused on in initial testing because of the large amount of available data, however eXpress can be easily applied to allele-specific expression, metagenomics, and many other active areas of research.
Does eXpress only work in streaming mode?
No, eXpress is also able to accept a SAM or BAM file as input (see Manual and Getting Started for more information on how to use different inputs. If given an alignment file, eXpress is also able to do multiple passes over the data to improve accuracy. See the Manual for more details on different iterative modes.
Does eXpress accept both single-end and paired-end read alignments?
Yes, eXpress accepts both types of reads, even in the same alignment file. However, to maximize accuracy of the bias correction and fragment length calculations, you should not mix reads generated in separate library preps. See below for more details.
Can I give eXpress a combined set of reads generated in different library preparations or with different read lengths?
Yes, but you have to do so carefully. eXpress can take a comma-separated list of SAM filenames as its command-line input. When this is done, eXpress will assume that each alignment file contains reads fir the same sample but sequenced from a different library preparation and/or using a different read length. It will then estimate auxiliary parameters independently for each set but will combine the count/abundance results. Since some overall power is lost when smaller alignment sets are used, any reads generated from the same library preparation using the same read lengths should be combined into a single alignment file. See the Manual for more details.
Why does eXpress say it processed 0 fragments when I gave it a bunch of paired-end reads?
This is most likely because the paired-end reads do not have matching names in your alignment file. eXpress knows that your reads are supposed to be paired-end in this case, but it assumes one end is unmapped and tosses the fragment when the matching reads cannot be found. See the Manual for details on how to address this issue.
Does eXpress support Bowtie2 mappings?
Yes, however it is highly recommended you limit the length of indels allowed to avoid unlikely mappings to splice variants. For example, we recommend you use the following settings: -a -X 600 --rdg 6,5 --rfg 6,5 --score-min L,-.6,-.4 --no-discordant --no-mixed
What do I do if I need help and my question isn't answered here?
First, you should check to see if you question is answered in the Manual or Tutorial. If not, you should see if it has been addressed on our User Forum or the SeqAnswers Forum. If you still can't find your answer, feel free to contact us at email@example.com. We are busy but will do our best to respond as soon as possible!